LLVM-Bench (Live-SWE-agent): leaderboard
LLVM-Bench (2026) poses 423 real, reproducible LLVM compiler issues; a patch counts as resolved only if it applies, builds and passes the full LLVM test suite with the issue tests in the LLVM-Gym environment. Resolved rate in percent under the self-evolving Live-SWE-agent scaffold with at most 50 turns, higher is better.
Metric: Resolved rate (%; share of the 423 validated LLVM issues, versions 18-21, whose generated patch applies, builds and passes the full LLVM test suite including the issue tests in LLVM-Gym; self-evolving Live-SWE-agent scaffold with at most 50 interaction turns, temperature 0, 8,192-token generation limit). Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 4 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Live-SWE-agent + deepseek-v3.2 | 8.51 | |
| 2 | Live-SWE-agent + grok-code-fast-1 | 7.57 | |
| 3 | Live-SWE-agent + gemini-3-flash | 4.73 | |
| 4 | Live-SWE-agent + qwen3-coder-plus | 2.6 |
Interactive version: theaggregate.ai/benchmark?slug=llvm-bench-live-swe-agent · How It Works · Data refreshed daily, snapshot 2026-11-13.